The present disclosure relates to machine vision, and more particularly to a computer vision system configured for saccadic vision.
Many machine learning approaches seek to recognize an object in an image by parallel processing of all pixels in the image. These methods can be computationally expensive.